Legal claims defining the scope of protection. Each claim is shown in both the original legal language and a plain English translation.
1. A computer-implemented method for reconfiguring data flow across network channels, at least a portion of the method being performed by a computing system comprising at least one computer processor, the method comprising: monitoring, in a replication environment, a first network channel and a second network channel that transmit replication data, where the first network channel transmits the replication data using a first network protocol and the second network channel transmits the replication data using a second network protocol that is different than the first network protocol; identifying one or more characteristics of the first and second network channels; obtaining one or more performance metrics of the first and second network channels at least in part by determining a network latency of both the first and second network channels; determining that the network latency of at least one of the first and second network channels exceeds a severity threshold; in response to determining that the network latency of at least one of the first and second network channels exceeds the severity threshold: identifying a round-trip delay time of both the first and second network channels; determining whether a difference between the round-trip delay time of the first network channel and the round-trip delay time of the second network channel is less than a predetermined amount; determining whether the network latency of at least one of the first and second network channels is caused by propagation or congestion based on whether the difference between the round-trip delay time of the first network channel and the round-trip delay time of the second network channel is less than the predetermined amount; reconfiguring data flow within the replication environment based on both the characteristics and the performance metrics of the first and second network channels, wherein reconfiguring the data flow based on the performance metrics comprises reconfiguring the data flow based at least in part on whether the network latency is cause by propagation or congestion.
A computer system monitors two network channels (Channel A and Channel B) used for replicating data. Channel A uses one network protocol, and Channel B uses a different network protocol. The system identifies characteristics (e.g., bandwidth, protocol type) and performance metrics (e.g., latency) for both channels. Specifically, it measures the network latency of each channel, and if the latency of either exceeds a threshold, the system measures the round-trip delay time for each channel. Based on whether the difference in round-trip times is above or below a defined amount, the system determines if latency is caused by propagation delay or network congestion. Data flow is then reconfigured based on both the identified characteristics and the determined cause of latency in each channel.
2. The computer-implemented method of claim 1 , wherein: the first network channel uses a Fibre channel protocol; the second network channel uses an Ethernet protocol.
The data replication system described previously uses Fibre Channel protocol for network channel A and Ethernet protocol for network channel B. The system dynamically switches or adjusts data flow between the Fibre Channel and Ethernet channels based on their performance characteristics and latency issues to ensure optimal data replication.
3. The computer-implemented method of claim 1 , wherein at least one of the first and second network channels represents an InfiniBand channel.
In the described data replication system, either the network channel A or network channel B, or potentially both, use the InfiniBand protocol. The system monitors and dynamically manages data flow, taking into account performance characteristics like latency, to ensure optimal data replication across heterogeneous network technologies including InfiniBand.
4. The computer-implemented method of claim 1 , wherein: determining whether the difference between the round-trip delay time of the first network channel and the round-trip delay time of the second network channel is less than the predetermined amount comprises determining that the difference between the round-trip delay time of the first network channel and the round-trip delay time of the second network channel is less than the predetermined amount; determining whether the network latency of at least one of the first and second network channels is caused by propagation or congestion comprises determining that the network latency of at least one of the first and second network channels is caused by propagation based on the determination that the difference between the round-trip delay time of the first network channel and the round-trip delay time of the second network channel is less than the predetermined amount.
In the data replication system, when the difference in round-trip delay times between network channel A and network channel B is below a specific threshold, the system determines that any latency issues are due to propagation delay (inherent delay due to distance or physical properties of the channel). This determination informs how the system reconfigures data flow to mitigate the impact of propagation delays.
5. The computer-implemented method of claim 4 , further comprising: determining, based on the identified characteristics, that batching data segments transmitted using the first network channel improves network latency above a threshold amount but that batching data segments transmitted using the second network channel does not improve network latency above the threshold amount; in response to determining that the network latency of at least one of the first and second network channels is caused by propagation and determining that batching data segments transmitted using the first network channel improves network latency above the threshold amount but that batching data segments transmitted using the second network channel does not improve network latency above the threshold amount, reconfiguring data flow by batching data segments transmitted using the first network channel while not batching data segments transmitted using the second network channel.
Continuing from the previous description, the data replication system determines that batching data segments for transmission on network channel A improves latency above a threshold. However, batching on network channel B does not provide the same latency improvement. If propagation delay is detected, the system reconfigures data flow to batch data segments on channel A but not on channel B, optimizing performance by tailoring the data transmission strategy to each channel's characteristics.
6. The computer-implemented method of claim 1 , wherein: determining whether the difference between the round-trip delay time of the first network channel and the round-trip delay time of the second network channel is less than the predetermined amount comprises determining that the difference between the round-trip delay time of the first network channel and the round-trip delay time of the second network channel is greater than the predetermined amount; determining whether the network latency of at least one of the first and second network channels is caused by propagation or congestion comprises determining that the network latency of at least one of the first and second network channels is caused by congestion based on the determination that the difference between the round-trip delay time of the first network channel and the round-trip delay time of the second network channel is greater than the predetermined amount.
In the data replication system, when the difference in round-trip delay times between network channel A and network channel B is above a specific threshold, the system determines that any latency issues are due to network congestion (e.g., network overload or bottlenecks). The system then uses this congestion determination to guide data flow reconfiguration strategies.
7. The computer-implemented method of claim 6 , wherein: determining that the network latency of at least one of the first and second network channels exceeds the severity threshold comprises determining that the network latency of the first network channel exceeds the severity threshold; determining that the network latency of at least one of the first and second network channels is caused by congestion comprises determining that the network latency of the first network channel is caused by congestion; the method further comprises removing the first network channel from a list of available data replication paths in response to determining that the network latency of the first network channel is caused by congestion.
In the data replication system, the system detects high latency on network channel A and determines, based on round-trip delay time differences, that the high latency on channel A is caused by network congestion. In response, the system removes channel A from the list of available data replication paths, preventing further data transmission via the congested channel until the congestion resolves, thus optimizing data flow across available channels.
8. The computer-implemented method of claim 1 , further comprising: after reconfiguring the data flow within the replication environment, monitoring the first and second network channels at a subsequent moment in time; determining an additional network latency of both the first and second network channels at the subsequent moment in time; determining that the additional network latency of both the first and second network channels does not exceed the severity threshold at the subsequent moment in time; reconfiguring the data flow within the replication environment at the subsequent moment in time according to a bandwidth of at least one of the first and second network channels based on determining that the additional network latency of at least one of the first and second network channels does not exceed the severity threshold at the subsequent moment in time.
After reconfiguring the data flow in the replication environment (e.g. switching data replication to Channel B due to Channel A congestion), the system continues to monitor network channels A and B. If, at a later time, latency on both channels is below the severity threshold, the system reconfigures the data flow based on the bandwidth of each channel. This allows the system to revert to using the channel with the highest bandwidth if latency is no longer a concern.
9. The computer-implemented method of claim 1 , wherein the replication environment replicates data using only native protocols for channels within the replication environment.
The data replication system performs data replication using only native network protocols within the replication environment, avoiding the use of encapsulation or tunneling technologies like Fibre Channel over Ethernet (FCoE). It directly utilizes the capabilities of protocols like Fibre Channel, Ethernet, and InfiniBand without adding protocol translation layers.
10. A system for reconfiguring data flow across network channels, the system comprising: a monitoring module, stored in memory, that monitors, in a replication environment, a first network channel and a second network channel that transmit replication data, where the first network channel transmits the replication data using a first network protocol and the second network channel transmits the replication data using a second network protocol that is different than the first network protocol; an identification module, stored in memory, that identifies one or more characteristics of the first and second network channels; an obtaining module, stored in memory, that obtains one or more performance metrics of the first and second network channels at least in part by determining a network latency of both the first and second network channels, wherein: the obtaining module further determines that the network latency of at least one of the first and second network channels exceeds a severity threshold; in response to determining that the network latency of at least one of the first and second network channels exceeds the severity threshold, the obtaining module: identifies a round-trip delay time of both the first and second network channels; determines whether a difference between the round-trip delay time of the first network channel and the round-trip delay time of the second network channel is less than a predetermined amount; determines whether the network latency of at least one of the first and second network channels is caused by propagation or congestion based on whether the difference between the round-trip delay time of the first network channel and the round-trip delay time of the second network channel is less than the predetermined amount; a reconfiguring module, stored in memory, that reconfigures data flow within the replication environment based on both the characteristics and the performance metrics of the first and second network channels, wherein the reconfiguring module reconfigures the data flow based on the performance metrics by reconfiguring the data flow based at least in part on whether the network latency is cause by propagation or congestion; at least one physical processor configured to execute the monitoring module, the identification module, the obtaining module, and the reconfiguring module.
A data replication system includes a monitoring module that monitors two network channels (Channel A and Channel B) used for replication with different protocols. An identification module determines characteristics of each channel. An obtaining module measures performance metrics, including latency, and determines if the latency exceeds a severity threshold. If exceeded, round-trip times are measured to determine if latency is due to propagation or congestion. A reconfiguring module then adjusts data flow based on channel characteristics and the cause of latency. The modules are executed by at least one physical processor.
11. The system of claim 10 , wherein: the first network channel uses a Fibre channel protocol; the second network channel use an Ethernet protocol.
In the previously described system, network channel A utilizes Fibre Channel protocol, while network channel B utilizes Ethernet protocol. The system monitors and dynamically manages the data flow between these two channel types, reconfiguring the transmission based on observed latency and performance metrics of each.
12. The system of claim 10 , wherein: the replication environment is associated with a backup server that facilitates transmitting data from a source computing device to a target computing device; the monitoring module, the identification module, the obtaining module, and the reconfiguring module operate as part of a replication channel controller; the reconfiguring module reconfigures data flow within the replication environment by instructing a transport adapter to reconfigure data flow between the source computing device and the target computing device; upon receiving the instructions from the reconfiguring module, the transport adapter transmits data using the first and second network channel in accordance with the instructions by hooking at least one data stream Application Programming Interface.
In the described system, the replication environment involves a backup server transferring data between a source and target device. The modules (monitoring, identification, obtaining, and reconfiguring) act as a replication channel controller. The reconfiguring module instructs a transport adapter to adjust the data flow between the source and target. The transport adapter "hooks" data stream APIs to transmit data on channels A and B according to those reconfiguration instructions.
13. The system of claim 10 , wherein at least one of the first and second network channels represents an InfiniBand channel.
Either network channel A or network channel B, or both, utilize InfiniBand protocol in the data replication system. The system is designed to monitor and dynamically manage data flow considering the performance characteristics of InfiniBand alongside other protocols like Ethernet or Fibre Channel.
14. The system of claim 10 , wherein: the obtaining module determines whether the difference between the round-trip delay time of the first network channel and the round-trip delay time of the second network channel is less than the predetermined amount by determining that the difference between the round-trip delay time of the first network channel and the round-trip delay time of the second network channel is less than the predetermined amount; the obtaining module determines whether the network latency of at least one of the first and second network channels is caused by propagation or congestion by determining that the network latency of at least one of the first and second network channels is caused by propagation based on the determination that the difference between the round-trip delay time of the first network channel and the round-trip delay time of the second network channel is less than the predetermined amount.
The system determines the cause of latency by calculating the difference in round-trip delay times. If the difference is below a threshold, the system determines that latency is caused by propagation delay (e.g., physical distance). The system adjusts data flow based on this determination.
15. The system of claim 14 , wherein: the reconfiguring module determines, based on the identified characteristics, that batching data segments transmitted using the first network channel improves network latency above a threshold amount but that batching data segments transmitted using the second network channel does not improve network latency above the threshold amount; in response to determining that the network latency of at least one of the first and second network channels is caused by propagation and determining that batching data segments transmitted using the first network channel improves network latency above the threshold amount but that batching data segments transmitted using the second network channel does not improve network latency above the threshold amount, the reconfiguring module reconfigures data flow by batching data segments transmitted using the first network channel while not batching data segments transmitted using the second network channel.
The system determines that batching data segments on network channel A improves latency above a threshold, but batching on network channel B does not. If latency is caused by propagation delay, the system reconfigures data flow to batch data segments on channel A but not on channel B, optimizing performance based on individual channel characteristics.
16. The system of claim 10 , wherein: the obtaining module determines whether the difference between the round-trip delay time of the first network channel and the round-trip delay time of the second network channel is less than the predetermined amount by determining that the difference between the round-trip delay time of the first network channel and the round-trip delay time of the second network channel is greater than the predetermined amount; the obtaining module determines whether the network latency of at least one of the first and second network channels is caused by propagation or congestion by determining that the network latency of at least one of the first and second network channels is caused by congestion based on the determination that the difference between the round-trip delay time of the first network channel and the round-trip delay time of the second network channel is greater than the predetermined amount.
The system determines latency cause based on round-trip time differences. If the difference in round-trip times between channel A and B is above a threshold, the system determines that latency is caused by network congestion. Data flow adjustments are then made to address the identified congestion issue.
17. The system of claim 16 , wherein: the obtaining module determines that the network latency of at least one of the first and second network channels exceeds the severity threshold by determining that the network latency of the first network channel exceeds the severity threshold; the obtaining module determines that the network latency of at least one of the first and second network channels is caused by congestion by determining that the network latency of the first network channel is caused by congestion; the reconfiguring module removes the first network channel from a list of available data replication paths in response to the determination that the network latency of the first network channel is caused by congestion.
The system detects high latency on network channel A and determines that this latency is due to network congestion. In response, the system removes network channel A from the list of available data replication paths, rerouting data traffic to other, less congested channels.
18. The system of claim 10 , wherein: the first network channel represents a dedicated resource used only for replication; the second network channel represents a resource used for both replication and other network communication.
In the data replication system, network channel A represents a dedicated resource exclusively used for data replication tasks. Network channel B represents a shared resource that is used for both data replication and other general network communications. The system manages data flow differently between the dedicated replication channel and the shared network resource based on their respective performance characteristics.
19. A non-transitory computer-readable medium comprising one or more computer-readable instructions that, when executed by at least one processor of a computing device, cause the computing device to: monitor, in a replication environment, a first network channel and a second network channel that transmit replication data, where the first network channel transmits the replication data using a first network protocol and the second network channel transmits the replication data using a second network protocol that is different than the first network protocol; identify one or more characteristics of the first and second network channels; obtain one or more performance metrics of the first and second network channels at least in part by determining a network latency of both the first and second network channels; determine that the network latency of at least one of the first and second network channels exceeds a severity threshold; in response to determining that the network latency of at least one of the first and second network channels exceeds the severity threshold: identifying a round-trip delay time of both the first and second network channels; determining whether a difference between the round-trip delay time of the first network channel and the round-trip delay time of the second network channel is less than a predetermined amount; determining whether the network latency of at least one of the first and second network channels is caused by propagation or congestion based on whether the difference between the round-trip delay time of the first network channel and the round-trip delay time of the second network channel is less than the predetermined amount; reconfigure data flow within the replication environment based on both the characteristics and the performance metrics of the first and second network channels, wherein reconfiguring the data flow based on the performance metrics comprises reconfiguring the data flow based at least in part on whether the network latency is cause by propagation or congestion.
A computer-readable medium contains instructions that, when executed, cause a computer to monitor two network channels (Channel A and Channel B) with different protocols, used for replicating data. The computer identifies channel characteristics and measures performance metrics, including latency. If latency exceeds a threshold, round-trip times are measured to determine if latency is due to propagation or congestion. Data flow is then reconfigured based on channel characteristics and the cause of latency.
20. The non-transitory computer-readable medium of claim 19 , wherein: the first network channel comprises a Fibre channel; the second network channel comprises an Ethernet channel; the replication environment implements both a Fibre Channel Protocol and an Ethernet Channel Protocol without using Fibre Channel over Ethernet.
The computer-readable medium for data replication from the previous description implements Fibre Channel on the first network channel and Ethernet on the second network channel. The replication environment directly implements both protocols *without* using Fibre Channel over Ethernet (FCoE) which encapsulates Fibre Channel within Ethernet. This approach enables native protocol usage with dynamic management of replication channels.
Unknown
September 26, 2017
Browse 5M+ US patents with plain-English claim translations and AI-generated analysis.